Search

In this paper, we present a novel vibrotactile rendering algorithm for producing real-time tactile interactions suitable for virtual reality applications. The algorithm uses an energy model to produce smooth tactile sensations by continuously recalculating the location of a phantom actuator that represents a virtual touch point. It also employs syncopations in its rendered amplitude to produce artificial perceptual anchors that make the rendered vibrotactile patterns more recognizable. We conducted two studies to compare this Syncopated Energy algorithm to a standard real-time Grid Region algorithm for rendering touch patterns at different vibration amplitudes and frequencies. We found that the Grid Region algorithm afforded better recognition, but that the Syncopated Energy algorithm was perceived to produce smoother patterns at higher amplitudes. Additionally, we found that higher amplitudes afforded better recognition while a moderate amplitude yielded more perceived continuity. We also found that a higher frequency resulted in better recognition for fine-grained tactile sensations and that frequency can affect perceived continuity.

Biometrics security is a dynamic research area spurred by the need to protect personal traits from threats like theft, non-authorised distribution, reuse and so on. A widely investigated solution to such threats consists of processing the biometric signals under encryption, in order to avoid any leakage of information towards non-authorised parties. In this study, the authors propose to leverage on the superior performance of multimodal biometric recognition to improve the efficiency of a biometric-based authentication protocol operating on encrypted data under the malicious security model. In the proposed protocol, authentication relies on both facial and iris biometrics, whose representation accuracy is specifically tailored to the trade-off between recognition accuracy and efficiency. From a cryptographic point of view, the protocol relies on Damgard et al. SPDZ. Experimental results show that the multimodal protocol is faster than corresponding unimodal protocols achieving the same accuracy.

In this paper, two approaches to building extraction from satellite imagery and height data obtained from stereo images or LIDAR are compared. The first approach consists of detecting high-rise objects in a digital surface model and then improving recognition accuracy using segmentation of spectral information. The second approach uses the U-Net convolutional neural network, which showed the best results for the extraction of objects from aerospace images on a number of large datasets. Extensive experiments were carried out to evaluate the dependence of the quality of U-Net-based building extraction on the different data types (including high-resolution satellite images and digital surface model data). Building extraction quality of the trained network was also evaluated on satellite images with different spatial resolutions. © 2018 CEUR-WS. All rights reserved.

The performance of automatic speech recognition based on coded-decoded speech heavily depends on the quality of the transmitted signals, determined by channel impairments. This paper examines relationships between speech recognition performance and measurements of speech quality and intelligibility over transmission channels. Different to previous studies, the effects of super-wideband transmissions are analyzed and compared to those of wideband and narrowband channels. Furthermore, intelligibility scores, gathered by conducting a listening test based on logatomes, are also considered for the prediction of automatic speech recognition results. The modern instrumental measurement techniques POLQA and POLQA-based intelligibility have been respectively applied to estimate the quality and the intelligibility of transmitted speech. Based on our results, polynomial models are proposed that permit the prediction of speech recognition accuracy from the subjective and instrumental measures, involving a number of channel distortions in the three bandwidths. This approach can save the costs of performing automatic speech recognition experiments and can be seen as a first step towards a useful tool for communication channel designers. Copyright © 2017 ISCA. Amazon Alexa; Apple; DiDi; et al.; Furhat Robotics; Microsoft

The performance of automatic speech recognition based on coded-decoded speech heavily depends on the quality of the transmitted signals, determined by channel impairments. This paper examines relationships between speech recognition performance and measurements of speech quality and intelligibility over transmission channels. Different to previous studies, the effects of super-wideband transmissions are analyzed and compared to those of wideband and narrowband channels. Furthermore, intelligibility scores, gathered by conducting a listening test based on logatomes, are also considered for the prediction of automatic speech recognition results. The modern instrumental measurement techniques POLQA and POLQA-based intelligibility have been respectively applied to estimate the quality and the intelligibility of transmitted speech. Based on our results, polynomial models are proposed that permit the prediction of speech recognition accuracy from the subjective and instrumental measures, involving a number of channel distortions in the three bandwidths. This approach can save the costs of performing automatic speech recognition experiments and can be seen as a first step towards a useful tool for communication channel designers. Copyright © 2017 ISCA. Amazon Alexa; Apple; DiDi; et al.; Furhat Robotics; Microsoft

Paper presents development of the acoustic model for Croatian language for automatic speech recognition (ASR). Continuous speech recognition is performed by means of the Hidden Markov Models (HMM) implemented in the HMM Toolkit (HTK). In order to adjust the HTK to the native language a novel algorithm for Croatian language transcription (CLT) has been developed. It is based on phonetic assimilation rules that are applied within uttered words. Phonetic questions for state tying of different triphone models have also been developed. The automated system for training and evaluation of acoustic models has been developed and integrated with the new graphical user interface (GUI). Targeted applications of this ASR system are stress inoculation training (SIT) and virtual reality exposure therapy (VRET). Adaptability of the model to a closed set of speakers is important for such applications and this paper investigates the applicability of the HTK tool for typical scenarios. Robustness of the tool to a new language was tested in matched conditions by a parallel training of an English model that was used as a baseline. Ten native Croatian speakers participated in experiments. Encouraging results were achieved and reported with the developed model for Croatian language.

With the rapid development of Internet-of-Things (IoT), face scrambling has been proposed for privacy protection during IoT-targeted image/video distribution. Consequently, in these IoT applications, biometric verification needs to be carried out in the scrambled domain, presenting significant challenges in face recognition. Since face models become chaotic signals after scrambling/encryption, a typical solution is to utilize the traditional data-driven face recognition algorithms. While chaotic pattern recognition is still a challenging task, in this paper, we propose a new ensemble approach - many-kernel random discriminant analysis (MK-RDA) - to discover discriminative patterns from the chaotic signals. We also incorporate a salience-aware strategy into the proposed ensemble method to handle the chaotic facial patterns in the scrambled domain, where the random selections of features are made on semantic components via salience modeling. In our experiments, the proposed MK-RDA was tested rigorously on three human face data sets: the ORL face data set, the PIE face data set, and the PUBFIG wild face data set. The experimental results successfully demonstrate that the proposed scheme can effectively handle the chaotic signals and significantly improve the recognition accuracy, making our method a promising candidate for secure biometric verification in the emerging IoT applications. 2016 IEEE. Scopus

The inherent skin patterns created at the joints in the finger exterior are referred as finger knuckle-print. It is exploited to identify a person in a unique manner because the finger knuckle print is greatly affluent in textures. In biometric system, the region of interest is utilized for the feature extraction algorithm. In this paper, local and global features are extracted separately. Fast Discrete Orthonormal Stockwell Transform is exploited to extract the local features. Global feature is attained by escalating the size of Fast Discrete Orthonormal Stockwell Transform to infinity. Two features are fused to increase the recognition accuracy. A matching distance is calculated for both the features individually. Then two distances are merged mutually to acquire the final matching distance. The proposed scheme gives the better performance in terms of equal error rate and correct recognition rate., {"references":["A. Kumar, and C. Ravikanth, \"Personal authentication using finger\nknuckle surface\", IEEE T INF FOREN SEC, vol. 4, no. 1, pp. 98-109,\n2009.","Z. Lin, Personal authentication using finger knuckle print, doctoral diss.,\nThe Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong\nKong, 2011.","Y. Wanknou, S. Changyin, and S. Zhongxi, \"Finger-Knuckle-Print\nRecognition Using Gabor Feature and OLDA\", in Proceedings of the\n30th Chinese Control Conference (CCC '11), pp. 2975-2978, Yantai,\nJuly 2011.","X. Jing, W. Li, C. Lan, Y. Yao, X. Cheng, and L. Han, \"Orthogonal\nComplex Locality Preserving Projections based on Image Space Metric\nfor Finger-Knuckle-Print Recognition\", in Proceedings of the\nInternational Conference on Hand-Based Biometrics (ICHB '11), pp. 1-\n6, Hong Kong, November 2011.","L. Zhang, L. Zhang, and D. Zhang, \"Monogenic Code: A Novel Fast\nFeature Coding Algorithm with Applications to Finger-Knuckle-Print\nRecognition\", in Proceedings of the International Workshop on\nEmerging Techniques and Challenges for Hand Based Biometrics\n(ETCHB 10'), 2010.","S. Neware, K. Mehta, and A.S. Zadgaonkar, \"Finger Knuckle\nIdentification using Principal Component Analysis and Nearest Mean\nClassifier\", International Journal of Computer Applications, vol. 70,\nno.9, pp. 18-23, 2013.","K. Ito, T. Aoki, H. Nakajima, K. Kobayashi, and T. Higuchi, \"A\npalmprint recognition algorithm using phase-only correlation\", IEICE T\nFUND ELECTR, vol. E91-A, no.4, pp.1023–1030, 2008.","K. Ito, H. Nakajima, K. Kobayashi, T. Aoki, and T.Higuchi, \"A\nfingerprint matching algorithm using phase-only correlation\", IEICE T\nFUND ELECTR, vol. E91-A, no.3, pp. 682–691, 2004.","\"Finger Knuckle-print Database\",\nhttp://www4.comp.polyu.edu.hk/~biometrics\n[10] W.K. Kong, and D. Zhang, \"Competitive coding scheme for palmprint\nverification\", in Proceedings of 17th International Conference on Pattern\nRecognition (ICPR '04), vol. 1, pp. 520–523, August 2004.\n[11] L.Zhang, L.Zhang, David Zhang, and H.Zhu, \"Ensemble of local and\nglobal information for finger-knuckle print recognition\", PATTERN\nRECOGN, vol. 44, pp. 1990-1998, 2011."]}

Proposed a novel extension to Histogram Equalization method for noise compensation.We perform a sub-band specific equalization on the noisy cepstral features.Histogram analysis and recognition results show usefulness of the proposed approach.Favorable in real-time systems due to superior performance and computational benifits. This paper describes a novel framework to sub-band based Histogram Equalization (HEQ) applied to robust speech recognition. We propose a frequency band specific equalization to compensate the noise distortion on the individual frequency bands. The proposed equalization framework is a two step process. In the first step, conventional histogram equalization is done. By analyzing the histograms of equalized cepstra, we show that the first stage of conventional HEQ approach does not compensate the sub-band specific noise distortion, even though the overall histogram is normalized. Hence, in the second stage, sub-band specific histogram equalization is done. Every frame of cepstral coefficients is decomposed into low-frequency (LF) cepstra and high-frequency (HF) cepstra. Separate equalization is done on LF and HF cepstra to compensate LF and HF specific noise distortion. The cepstra corresponding to the LF and HF bands are obtained by using simple averaging and differencing filters on the cepstral components within a particular frame. The proposed approach is referred to as Sub-band Histogram Equalization (S-HEQ). Using histogram analysis, we show that the S-HEQ approach is able to compensate for the sub-band specific noise distortion. S-HEQ approach shows a consistent improvement over the conventional HEQ approach with a relative improvement of 12 % and 22.10 % over conventional HEQ in WER on Aurora-2 and Aurora-4 databases respectively. Proposed equalization approach can also be used with the deep neural network based systems and has shown a consistent improvement in the recognition accuracies over conventional HEQ. Finally, the efficacy of the proposed S-HEQ approach for embedded real-time speech applications is shown by comparing the performance and computational complexity trade-off with other state-of-the-art noise compensation methods.

Dan je uvod u osnove tehnologije za automatsko raspoznavanje govora. Među navedenim ubrajaju se: praktične primjene sustava za automatsko raspoznavanje govora; prepreke i izazovi pri dizajniranju kvalitetnog sustava; princip rada i uloga pojedinih dijelova sustava; glavne karakteristike i metrike ASR sustava; algoritmi, tehnike i modeli korišteni kod automatskog raspoznavanja govora; te povijesni pregled razvoja ASR tehnologija. Prikazane su neke nove tehnologije koje još nisu implementirane u postojeće komercijalne programe, a znatno povećavaju točnost prepoznavanja riječi u govoru. Istražene su mogućnosti današnjih komercijalnih programa i maksimalna točnost prepoznavanja koju mogu ostvariti, koja se pokazala vrlo visokom (preko 90%). Prikazan je postupak izrade programa za prepoznavanje govora, konfiguracije ulaza, te finog podešavanja svih potrebnih postavki operativnog sustava. Izrađeni program podvrgnut je treniranju i detaljnom testiranju, kako bi se ustanovilo koji sve elementi i u kolikoj mjeri utječu na točnost prepoznavanja govora, te koja je najveća točnost prepoznavanja koju pritom može ostvariti. Doneseni su čvrsti zaključci i dana je osnova na temelju koje se može lakše razumjeti i promatrati daljnji razvoj tehnologija za automatsko raspoznavanje govora. Basics of speech recognition technology are introduced. Among the mentioned topics are: practical uses of speech recognition system; obstacles and challenges in quality system design; principle of operation and specific part roles; main characteristics and metrics of ASR system; algorithms, techniques and models used in automatic speech recognition; and historical overview of ASR technology development. This paper gives insight into some of the new technologies which have the ability of substantially increasing speech recognition accuracy, but have not yet been implemented into existing commercial programs. Abilities of today's commercial programs were explored, as well as maximum speech recognition accuracy they could achieve, which was very high by the way (over 90%). This paper also displays the procedure of making a speech recognition application, configuring input and fine tuning operating system settings. The application was trained and thoroughly tested in order to find out which elements affect recognition accuracy and in what measure, as well as what's the highest accuracy the application can accomplish. Solid conclusions are made and basis is obtained upon which one can easier understand and observe further speech recognition technology development.

Date of Conference: 23-25 April 2014 Conference name: 22nd Signal Processing and Communications Applications Conference (SIU) 2014 Bu makalede, iki diferansiyel kızılberisi algılayıcı (PIR) ve bir kamera kullanılarak geliştirilen el jestleri algılama ve sınıflandırma sistemi tanıtılmaktadır. İzlenen alanda diferansiyel PIR algılayıcı dizisi ile hareket varlığı araştırılır. Bir hareket algılanması durumunda kamera yardımıyla söz konusu hareketin el olup olmadığına, el ise çok modlu sistem verilerinin birlikte değerlendirilmesiyle hareketin hangi tanımlı sınıfa ait olduğuna karar verilir. Kamera ile el jestleri algılama ve hareketleri sınıflandırma aşamasında ten algılama ve dışbükey zarf-gedik hesaplama yöntemleri kullanılır. Farklı el hareketlerinin PIR algılayıcı verileri yardımıyla sınıflandırılması Winner-Take-All (WTA) imza metoduyla gerçekleştirilir. Bu makalenin temel katkısı, WTA imza kodlarının tek boyutlu sinyallerin sınıflandırılmasında kullanılabileceğini ve çoklu algılayıcı tümleştirmesiyle jestleri tanıma sonuçlarının geliştirilebileceğini göstermektir. In this paper, a hand gesture detection and classification system using two differential Pyro-electric Infrared (PIR) sensors and a camera is introduced. Motion presence is investigated in the area of interest using two differential PIR sensors. In the case of any motion detection, a decision whether it belongs to a hand or not is made by using camera and if it is a hand, which gesture belongs to which predefined class is determined by evaluating each system data together. In the stage of hand gesture detection and classification using camera, skin detection and convex hull-defect algorithms are used. Classification of different hand gestures by using differential PIR sensors is carried out by Winner-Take-All (WTA) hash method. The main contribution of this paper is to show that WTA hash codes can be utilized in classification of 1-D signals and gesture recognition accuracy can be improved by multi-sensor fusion. © 2014 IEEE.

Duration of training/test data has a considerable effect on the performance of a speaker recognition system. In this paper, we analyze the effect of training and test data duration and speaker gender on the performance of speaker recognition systems. Gaussian mixture models-universal background model (GMM-UBM), vector quantization-universal background model (VQ-UBM), support vector machines-generalized linear discriminant sequence kernel (SVM-GLDS) and support vector machines with GMM supervectors (GSV-SVM) are the classifiers we use for speaker recognition. Experimental results conducted on NIST 2002 and NIST 2005 speaker recognition evaluation (SRE) databases show that recognition performance breaks down when short utterances are used for training and testing independent from the recognizer (e.g. equal error rate (EER) reduces from 10.33% to 27.86% on NIST 2005) and GSV-SVM system yields higher EER than other methods in the case of using short utterances. It is also shown that recognition accuracy for male speakers are higher than female independent from database and classifier.

Over the last decade, an increasing number of studies have focused on automated recognition of human emotions by machines. However, performances of machine emotion recognition studies are difficult to interpret because benchmarks have not been established. To provide such a benchmark, we compared machine with human emotion recognition. We gathered facial expressions, speech, and physiological signals from 17 individuals expressing 5 different emotional states. Support vector machines achieved an 82% recognition accuracy based on physiological and facial features. In experiments with 75 humans on the same data, a maximum recognition accuracy of 62.8% was obtained. As machines outperformed humans, automated emotion recognition might be ready to be tested in more practical applications.

Bu çalışma, 25-29 Ağustos 2013 tarihleri arasında Lyon[Fransa]’da düzenlenen 14. Annual Conference of the International-Speech-Communication-Association (INTERSPEECH 2013)’da bildiri olarak sunulmuştur. We study the problem of vocal effort mismatch in speaker verification. Changes in speaker's vocal effort induce changes in fundamental frequency (F0) and formant structure which introduce unwanted intra-speaker variations to features. We compare seven alternative spectrum estimators in the context of melfrequency cepstral coefficient (MFCC) extraction for speaker verification. The compared variants include traditional FFT spectrum and six parametric all-pole models. Experimental results on the NIST 2010 speaker recognition evaluation (SRE) corpus utilizing both GMM-UBM and more recent GMM supervector classifier indicate that spectrum estimation has a considerable impact on speaker verification accuracy under mismatched vocal effort conditions. The highest recognition accuracy was achieved using a particular variant of temporally weighted all-pole model, stabilized weighted linear prediction (SWLP). Academy of Finland (253120) Int Speech Commun Assoc Europa org Amazon Microsoft Google TcL SYTRAL European Language Resources Assoc Ouaero Imaginove VOCAPIA res Acapela Speech ocean ALDEBARAN Orange Vecsys IBM Res Raytheon BBN Technol Voxygen

Bu çalışma, 25-29 Ağustos 2013 tarihlerinde Lyon[Fransa]'da düzenlenen 14. Annual Conference of the International Speech Communication Association [Interspeech 2013]'da bildiri olarak sunulmuştur. Popular features for speech processing, such as mel-frequency cepstral coefficients (MFCCs), are derived from the short-term magnitude spectrum, whereas the phase spectrum remains unused. While the common argument to use only the magnitude spectrum is that the human ear is phase-deaf, phase-based features have remained less explored due to additional signal processing difficulties they introduce. A useful representation of the phase is the group delay function, but its robust computation remains difficult. This paper advocates the use of group delay functions derived from parametric all-pole models instead of their direct computation from the discrete Fourier transform. Using a subset of the vocal effort data in the NIST 2010 speaker recognition evaluation (SRE) corpus, we show that group delay features derived via parametric all-pole models improve recognition accuracy, especially under high vocal effort. Additionally, the group delay features provide comparable or improved accuracy over conventional magnitude-based MFCC features. Thus, the use of group delay functions derived from all-pole models provide an effective way to utilize information from the phase spectrum of speech signals. Academy of Finland (253120) Int Speech Commun Association Amazon Microsoft Google TcL SYTRAL European Language Resources Association Ouaero Imaginove VOCAPIA Research Acapela Speech Ocean ALDEBARAN Orange Vecsys IBM Research Raytheon BBN Technology Voxygen

Bu çalışma, 26-31 Mayıs 2013 tarihleri arasında Vancouver[Kanada]’da düzenlenen IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)’da bildiri olarak sunulmuştur. Text-independent speaker identification is studied using neutral and shouted speech in Finnish to analyze the effect of vocal mode mismatch between training and test utterances. Standard mel-frequency cepstral coefficient (MFCC) features with Gaussian mixture model (GMM) recognizer are used for speaker identification. The results indicate that speaker identification accuracy reduces from perfect (100 %) to 8.71 % under vocal mode mismatch. Because of this dramatic degradation in recognition accuracy, we propose to use a joint density GMM mapping technique for compensating the MFCC features. This mapping is trained on a disjoint emotional speech corpus to create a completely speaker- and speech mode independent emotion-neutralizing mapping. As a result of the compensation, the 8.71 % identification accuracy increases to 32.00 % without degrading the non-mismatched train-test conditions much. Inst Elect & Elect Engineers Inst Elect & Elect Engineers Signal Proc Soc

Item does not contain fulltext Over the last decade, an increasing number of studies have focused on automated recognition of human emotions by machines. However, performances of machine emotion recognition studies are difficult to interpret because benchmarks have not been established. To provide such a benchmark, we compared machine with human emotion recognition. We gathered facial expressions, speech, and physiological signals from 17 individuals expressing 5 different emotional states. Support vector machines achieved an 82% recognition accuracy based on physiological and facial features. In experiments with 75 humans on the same data, a maximum recognition accuracy of 62.8% was obtained. As machines outperformed humans, automated emotion recognition might be ready to be tested in more practical applications. 39 p.

Automatic recognition of suspects from forensic sketches is of considerable interest to the law enforcement agencies. However, this task is complex due to the heterogenous nature of face sketches and photographs. To address this challenge, previous approaches generally learn a transformation of a sketch to photo or a photo to sketch at the image or feature level in order to reduce the modality gap. Such a transformation may be indeterministic and if learned from training data, is likely to over-fit the sketch artist's drawing technique. Instead, we formulate the problem in the context of matching local self similarities which are independently computed from a face sketch and a photo. The proposed Facial Self Similarity (FSS) descriptor is obtained by correlation of a small face patch with its local neighborhood. Thus, our approach avoids the need of a modality transformation, while implicitly reducing the inter-modality gap. The proposed FSS descriptor is evaluated on the CUHK Face Sketch database using sketch-photo pairs of 311 subjects. The FSS descriptor demonstrates high recognition accuracy of 99.53% and outperforms current techniques. We also evaluate the robustness of the descriptor to anomalies such as matching sketches to blurred photographs. Refereed/Peer-reviewed

There is a great interest in dimensionality reduction techniques for tackling the problem of high-dimensional pattern classification. This paper addresses the topic of supervised learning of a linear dimension reduction mapping suitable for classification problems. The proposed optimization procedure is based on minimizing an estimation of the nearest neighbor classifier error probability, and it learns a linear projection and a small set of prototypes that support the class boundaries. The learned classifier has the property of being very computationally efficient, making the classification much faster than state-of-the-art classifiers, such as SVMs, while having competitive recognition accuracy. The approach has been assessed through a series of experiments, showing a uniformly good behavior, and competitive compared with some recently proposed supervised dimensionality reduction techniques. © 2010 Elsevier B.V. All rights reserved., Work partially supported by the Spanish projects TIN2008-04571 and Consolider Ingenio 2010: MIPRCV (CSD2007-00018).

Bu çalışma, 29 Ağustos-2 Eylül 2011 tarihleri arasında Barselona[İspanya]'da düzenlenen 19. European Signal Processing Conference (Eusipco-2011)'de bildiri olarak sunulmuştur. The universal background model (UBM) based classifiers have recently been popular for speaker recognition. In this paper, we propose a dimension reduction method using local principal component analysis to improve the performance of speaker verification systems, where maximum a Posteriori (MAP) adapted vector quantization classifier (VQ-MAP or VQ-UBM) is employed. The proposed system first partitions the UBM data into disjoint regions (clusters) via conventional VQ algorithm and PCA is performed on the set of feature vectors in each region to obtain transformation matrix. Then, multiple speaker model is constructed using the set of transformed feature vectors closest to each cluster through MAP adaptation. Conducting experiments on NIST 2001 SRE, it is shown that transforming the data onto a lower dimensional space by the proposed method improves the recognition accuracy.

Psichofizikiniame eksperimente su žmogumi tirta simetrijos įtaka skirtingo elementų skaičiaus figūroms atpažinti maskavimo, t. y. analizės laiko trūkumo, sąlygomis. Sunkiai verbalizuojamos testo figūros skyrėsi vertikalių ir horizontalių linijų skaičiumi ir simetrijos rodikliu. Testo figūrų rinkinius sudarė po lygiai simetrinių ir nesimetrinių figūrų, turinčių 4, 6 arba 8 linijos atkarpas. Stimulai buvo pateikiami šitokia tvarka: garsinis signalas, testo figūra (10 ms), tarpstimulinis intervalas, maskuojantis vaizdas, kurio pagrindu sudarytos testo figūros (500 ms). Tarpstimulinis intervalas parenkamas individualiai kiekvienam tiriamajam toks, kad teisingų atsakymų būtų 50-90 proc., ir bandymo metu šis intervalas nekeičiamas. Nustatyta, kad simetrinių keturių ir šešių linijos atkarpų figūrų atpažinimo tikslumas mažesnis už tokį pat atkarpų skaičių turinčių nesimetrinių figūrų atpažinimo tikslumą. Simetrijos įtaka aštuonių linijos atkarpų figūroms buvo priešinga: šios klasės simetrinės figūros buvo atpažįstamos geriau už nesimetrines figūras. Rezultatai interpretuojami, remiantis esamu optimalaus figūrų elementų skaičiaus ir simetrijos santykiu, sąlygojančiu efektyviausią atpažinimą. THE EFFECT OF SYMMETRY ON THE RECOGNITION ACCURACY OF PATTERNS OF DIFFERENT COMPLEXITY Ona Gurčinienė, Alvydas Šoliūnas Summary Symmetry is one of the most important determinants of visual perception. The influence of pattern symmetry on their recognition accuracy was investigated in the psychophysical experiment on human subjects under the conditions of backward masking. A rectangular frame of twelve line segments of equal length was used as a masking pattern (MP). The test patterns (TPs), nonverbal figures were the parts of MP composed by joining four, six, or eight line segments. Five figure sets consisting of equal number of symmetrical and asymmetrical figures of different complexity were used in the experiment. TP was presented for 10 ms, followed by MP for 500 ms with interstimulus interval, which was established for each subject individually as the shortest interval permitting 50-90 proc. recognition accuracy. Each subject took part in four to six test trial sessions. In each trial session, the equal number of symmetrical and asymmetrical figures composed by the same number of line segments was presented for recognition. Experimental data have demonstrated that effect of symmetry on pattern recognition depends on the number of pattern components. The asymmetrical figures consisting of four and six line segments were recognised more accurately than symmetrical ones, whereas the asymmetrical figures consisting of eight line segments have had the lower recognition accuracy than symmetrical ones. Experimental results were interpreted on the basis of optimal relation between the number of elements and symmetry that enables the greatest recognition probability of visual patterns.